Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2505.20137 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866910008053071872 |
|---|---|
| author | Goemaere, Cédric Oliviers, Gaspard Bogacz, Rafal Demeester, Thomas |
| author_facet | Goemaere, Cédric Oliviers, Gaspard Bogacz, Rafal Demeester, Thomas |
| contents | Predictive Coding (PC) offers a brain-inspired alternative to backpropagation for neural network training, described as a physical system minimizing its internal energy. However, in practice, PC is predominantly digitally simulated, requiring excessive amounts of compute while struggling to scale to deeper architectures. This paper reformulates PC to overcome this hardware-algorithm mismatch. First, we uncover how the canonical state-based formulation of PC (sPC) is, by design, deeply inefficient in digital simulation, inevitably resulting in exponential signal decay that stalls the entire minimization process. Then, to overcome this fundamental limitation, we introduce error-based PC (ePC), a novel reparameterization of PC which does not suffer from signal decay. Though no longer biologically plausible, ePC numerically computes exact PC weights gradients and runs orders of magnitude faster than sPC. Experiments across multiple architectures and datasets demonstrate that ePC matches backpropagation's performance even for deeper models where sPC struggles. Besides practical improvements, our work provides theoretical insight into PC dynamics and establishes a foundation for scaling PC-based learning to deeper architectures on digital hardware and beyond. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2505_20137 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | ePC: Fast and Deep Predictive Coding for Digital Hardware Goemaere, Cédric Oliviers, Gaspard Bogacz, Rafal Demeester, Thomas Machine Learning Artificial Intelligence Predictive Coding (PC) offers a brain-inspired alternative to backpropagation for neural network training, described as a physical system minimizing its internal energy. However, in practice, PC is predominantly digitally simulated, requiring excessive amounts of compute while struggling to scale to deeper architectures. This paper reformulates PC to overcome this hardware-algorithm mismatch. First, we uncover how the canonical state-based formulation of PC (sPC) is, by design, deeply inefficient in digital simulation, inevitably resulting in exponential signal decay that stalls the entire minimization process. Then, to overcome this fundamental limitation, we introduce error-based PC (ePC), a novel reparameterization of PC which does not suffer from signal decay. Though no longer biologically plausible, ePC numerically computes exact PC weights gradients and runs orders of magnitude faster than sPC. Experiments across multiple architectures and datasets demonstrate that ePC matches backpropagation's performance even for deeper models where sPC struggles. Besides practical improvements, our work provides theoretical insight into PC dynamics and establishes a foundation for scaling PC-based learning to deeper architectures on digital hardware and beyond. |
| title | ePC: Fast and Deep Predictive Coding for Digital Hardware |
| topic | Machine Learning Artificial Intelligence |
| url | https://arxiv.org/abs/2505.20137 |